Excess Winter Mortality (EWM) as a Dynamic Forensic Tool: Where, When, Which Conditions, Gender, Ethnicity and Age

Abstract

To investigate the dynamic issues behind intra- and international variation in EWM (Excess Winter Mortality) using a rolling monthly EWM calculation. This is used to reveal seasonal changes in the EWM calculation and is especially relevant nearer to the equator where EWM does not reach a peak at the same time each year. In addition to latitude country specific factors determine EWM. Females generally show higher EWM. Differences between the genders are highly significant and seem to vary according to the mix of variables active each winter. The EWM for respiratory conditions in England and Wales ranges from 44% to 83%, which is about double the all-cause mortality equivalent. A similar magnitude of respiratory EWM is observed in other temperate countries. Even higher EWM can be seen for specific respiratory conditions. Age has a profound effect on EWM with a peak at puberty and then increases EWM at older ages. The gap between male and female EWM seems to act as a diagnostic tool reflecting the infectious/metrological mix in each winter. Difference due to ethnicity are also observed. An EWM equivalent calculation for sickness absence demonstrates how other health-related variables can be linked to EWM. Midway between the equator and the poles show the highest EWM since such areas tend to neglect the importance of keeping dwellings warm in the winter. Pandemic influenza does not elevate EWM, although seasonal influenza plays a part each winter. Pandemic influenza and changes in influenza strain/variant mix do, however, create structural breaks in the time series and this implies that comparing EWM between studies conducted over different times can be problematic. Cancer is an excellent example of the usefulness of rolling method since cancer EWM drifts each year, in some years increasing winter EWM and in other years diminishing it. In addition, analysis of sub-national EWM in the UK reveals high spatiotemporal granularity indicating roles for infectious outbreaks. The rolling method gives greater insight into the dynamic nature of EWM, which otherwise lies concealed in the current static method.